* Author: Carl Worth <cworth@cworth.org>
*/
-#include "notmuch-private.h"
-
-#include <stdio.h>
-#include <errno.h>
-#include <time.h>
-#include <unistd.h>
-#include <sys/types.h>
-#include <sys/stat.h>
-#include <sys/time.h>
+#include "database-private.h"
#include <iostream>
-#include <gmime/gmime.h>
-
#include <xapian.h>
-using namespace std;
+#include <glib.h> /* g_strdup_printf, g_free, GPtrArray, GHashTable */
-struct _notmuch_database {
- char *path;
- Xapian::WritableDatabase *xapian_db;
- Xapian::TermGenerator *term_gen;
-};
+using namespace std;
#define ARRAY_SIZE(arr) (sizeof (arr) / sizeof (arr[0]))
-/* Xapian complains if we provide a term longer than this. */
-#define NOTMUCH_MAX_TERM 245
-
-/* These prefix values are specifically chosen to be compatible
- * with sup, (http://sup.rubyforge.org), written by
- * William Morgan <wmorgan-sup@masanjin.net>, and released
- * under the GNU GPL v2.
- */
-
typedef struct {
const char *name;
const char *prefix;
} prefix_t;
-prefix_t NORMAL_PREFIX[] = {
- { "subject", "S" },
- { "body", "B" },
- { "from_name", "FN" },
- { "to_name", "TN" },
- { "name", "N" },
- { "attachment", "A" }
+/* Here's the current schema for our database:
+ *
+ * We currently have two different types of documents: mail and timestamps.
+ *
+ * Mail document
+ * -------------
+ * A mail document is associated with a particular email message file
+ * on disk. It is indexed with the following prefixed terms:
+ *
+ * Single terms of given prefix:
+ *
+ * type: mail
+ *
+ * id: Unique ID of mail, (from Message-ID header or generated
+ * as "notmuch-sha1-<sha1_sum_of_entire_file>.
+ *
+ * thread: The ID of the thread to which the mail belongs
+ *
+ * Multiple terms of given prefix:
+ *
+ * ref: All unresolved message IDs from In-Reply-To and
+ * References headers in the message. (Once a referenced
+ * message is added to the database and the thread IDs
+ * are linked the corresponding "ref" term is dropped
+ * from the message document.)
+ *
+ * tag: Any tags associated with this message by the user.
+ *
+ * A mail document also has two values:
+ *
+ * TIMESTAMP: The time_t value corresponding to the message's
+ * Date header.
+ *
+ * MESSAGE_ID: The unique ID of the mail mess (see "id" above)
+ *
+ * Timestamp document
+ * ------------------
+ * A timestamp document is used by a client of the notmuch library to
+ * maintain data necessary to allow for efficient polling of mail
+ * directories. The notmuch library does no interpretation of
+ * timestamps, but merely allows the user to store and retrieve
+ * timestamps as name/value pairs.
+ *
+ * The timestamp document is indexed with a single prefixed term:
+ *
+ * timestamp: The user's key value (likely a directory name)
+ *
+ * and has a single value:
+ *
+ * TIMETAMPS: The time_t value from the user.
+ */
+
+/* With these prefix values we follow the conventions published here:
+ *
+ * http://xapian.org/docs/omega/termprefixes.html
+ *
+ * as much as makes sense. Note that I took some liberty in matching
+ * the reserved prefix values to notmuch concepts, (for example, 'G'
+ * is documented as "newsGroup (or similar entity - e.g. a web forum
+ * name)", for which I think the thread is the closest analogue in
+ * notmuch. This in spite of the fact that we will eventually be
+ * storing mailing-list messages where 'G' for "mailing list name"
+ * might be even a closer analogue. I'm treating the single-character
+ * prefixes preferentially for core notmuch concepts (which will be
+ * nearly universal to all mail messages).
+ */
+
+prefix_t BOOLEAN_PREFIX_INTERNAL[] = {
+ { "type", "T" },
+ { "thread", "G" },
+ { "ref", "XREFERENCE" },
+ { "timestamp", "XTIMESTAMP" },
};
-prefix_t BOOLEAN_PREFIX[] = {
- { "type", "K" },
- { "from_email", "FE" },
- { "to_email", "TE" },
- { "email", "E" },
- { "date", "D" },
- { "label", "L" },
- { "source_id", "I" },
- { "attachment_extension", "O" },
- { "msgid", "Q" },
- { "thread", "H" },
- { "ref", "R" }
+prefix_t BOOLEAN_PREFIX_EXTERNAL[] = {
+ { "tag", "K" },
+ { "id", "Q" }
};
-/* Similarly, these value numbers are also chosen to be sup
- * compatible. */
+int
+_internal_error (const char *format, ...)
+{
+ va_list va_args;
+
+ va_start (va_args, format);
-typedef enum {
- NOTMUCH_VALUE_MESSAGE_ID = 0,
- NOTMUCH_VALUE_THREAD = 1,
- NOTMUCH_VALUE_DATE = 2
-} notmuch_value_t;
+ vfprintf (stderr, format, va_args);
-static const char *
-find_prefix (const char *name)
+ exit (1);
+
+ return 1;
+}
+
+const char *
+_find_prefix (const char *name)
{
unsigned int i;
- for (i = 0; i < ARRAY_SIZE (NORMAL_PREFIX); i++)
- if (strcmp (name, NORMAL_PREFIX[i].name) == 0)
- return NORMAL_PREFIX[i].prefix;
+ for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_INTERNAL); i++)
+ if (strcmp (name, BOOLEAN_PREFIX_INTERNAL[i].name) == 0)
+ return BOOLEAN_PREFIX_INTERNAL[i].prefix;
+
+ for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++)
+ if (strcmp (name, BOOLEAN_PREFIX_EXTERNAL[i].name) == 0)
+ return BOOLEAN_PREFIX_EXTERNAL[i].prefix;
- for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX); i++)
- if (strcmp (name, BOOLEAN_PREFIX[i].name) == 0)
- return BOOLEAN_PREFIX[i].prefix;
+ INTERNAL_ERROR ("No prefix exists for '%s'\n", name);
return "";
}
-/* "128 bits of thread-id ought to be enough for anybody" */
-#define NOTMUCH_THREAD_ID_BITS 128
-#define NOTMUCH_THREAD_ID_DIGITS (NOTMUCH_THREAD_ID_BITS / 4)
-typedef struct _thread_id {
- char str[NOTMUCH_THREAD_ID_DIGITS + 1];
-} thread_id_t;
-
-static void
-thread_id_generate (thread_id_t *thread_id)
+const char *
+notmuch_status_to_string (notmuch_status_t status)
{
- static int seeded = 0;
- FILE *dev_random;
- uint32_t value;
- char *s;
- int i;
-
- if (! seeded) {
- dev_random = fopen ("/dev/random", "r");
- if (dev_random == NULL) {
- srand (time (NULL));
- } else {
- fread ((void *) &value, sizeof (value), 1, dev_random);
- srand (value);
- fclose (dev_random);
- }
- seeded = 1;
- }
-
- s = thread_id->str;
- for (i = 0; i < NOTMUCH_THREAD_ID_DIGITS; i += 8) {
- value = rand ();
- sprintf (s, "%08x", value);
- s += 8;
+ switch (status) {
+ case NOTMUCH_STATUS_SUCCESS:
+ return "No error occurred";
+ case NOTMUCH_STATUS_XAPIAN_EXCEPTION:
+ return "A Xapian exception occurred";
+ case NOTMUCH_STATUS_FILE_ERROR:
+ return "Something went wrong trying to read or write a file";
+ case NOTMUCH_STATUS_FILE_NOT_EMAIL:
+ return "File is not an email";
+ case NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID:
+ return "Message ID is identical to a message in database";
+ case NOTMUCH_STATUS_NULL_POINTER:
+ return "Erroneous NULL pointer";
+ case NOTMUCH_STATUS_TAG_TOO_LONG:
+ return "Tag value is too long (exceeds NOTMUCH_TAG_MAX)";
+ default:
+ case NOTMUCH_STATUS_LAST_STATUS:
+ return "Unknown error status value";
}
}
+/* XXX: We should drop this function and convert all callers to call
+ * _notmuch_message_add_term instead. */
static void
add_term (Xapian::Document doc,
const char *prefix_name,
if (value == NULL)
return;
- prefix = find_prefix (prefix_name);
+ prefix = _find_prefix (prefix_name);
term = g_strdup_printf ("%s%s", prefix, value);
- if (strlen (term) <= NOTMUCH_MAX_TERM)
+ if (strlen (term) <= NOTMUCH_TERM_MAX)
doc.add_term (term);
g_free (term);
}
static void
-find_messages_by_term (Xapian::Database *db,
- const char *prefix_name,
- const char *value,
- Xapian::PostingIterator *begin,
- Xapian::PostingIterator *end)
+find_doc_ids (notmuch_database_t *notmuch,
+ const char *prefix_name,
+ const char *value,
+ Xapian::PostingIterator *begin,
+ Xapian::PostingIterator *end)
{
Xapian::PostingIterator i;
char *term;
- term = g_strdup_printf ("%s%s", find_prefix (prefix_name), value);
+ term = g_strdup_printf ("%s%s", _find_prefix (prefix_name), value);
- *begin = db->postlist_begin (term);
+ *begin = notmuch->xapian_db->postlist_begin (term);
- if (end)
- *end = db->postlist_end (term);
+ *end = notmuch->xapian_db->postlist_end (term);
free (term);
}
-Xapian::Document
-find_message_by_docid (Xapian::Database *db, Xapian::docid docid)
+static notmuch_private_status_t
+find_unique_doc_id (notmuch_database_t *notmuch,
+ const char *prefix_name,
+ const char *value,
+ unsigned int *doc_id)
{
- return db->get_document (docid);
+ Xapian::PostingIterator i, end;
+
+ find_doc_ids (notmuch, prefix_name, value, &i, &end);
+
+ if (i == end) {
+ *doc_id = 0;
+ return NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND;
+ } else {
+ *doc_id = *i;
+ return NOTMUCH_PRIVATE_STATUS_SUCCESS;
+ }
}
-Xapian::Document
-find_message_by_message_id (Xapian::Database *db, const char *message_id)
+static Xapian::Document
+find_document_for_doc_id (notmuch_database_t *notmuch, unsigned doc_id)
{
- Xapian::PostingIterator i, end;
+ return notmuch->xapian_db->get_document (doc_id);
+}
- find_messages_by_term (db, "msgid", message_id, &i, &end);
+static notmuch_private_status_t
+find_unique_document (notmuch_database_t *notmuch,
+ const char *prefix_name,
+ const char *value,
+ Xapian::Document *document,
+ unsigned int *doc_id)
+{
+ notmuch_private_status_t status;
- if (i != end)
- return find_message_by_docid (db, *i);
- else
- return Xapian::Document ();
+ status = find_unique_doc_id (notmuch, prefix_name, value, doc_id);
+
+ if (status) {
+ *document = Xapian::Document ();
+ return status;
+ }
+
+ *document = find_document_for_doc_id (notmuch, *doc_id);
+ return NOTMUCH_PRIVATE_STATUS_SUCCESS;
}
+notmuch_message_t *
+notmuch_database_find_message (notmuch_database_t *notmuch,
+ const char *message_id)
+{
+ notmuch_private_status_t status;
+ unsigned int doc_id;
+
+ status = find_unique_doc_id (notmuch, "id", message_id, &doc_id);
+
+ if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND)
+ return NULL;
+
+ return _notmuch_message_create (notmuch, notmuch, doc_id, NULL);
+}
+
+/* Advance 'str' past any whitespace or RFC 822 comments. A comment is
+ * a (potentially nested) parenthesized sequence with '\' used to
+ * escape any character (including parentheses).
+ *
+ * If the sequence to be skipped continues to the end of the string,
+ * then 'str' will be left pointing at the final terminating '\0'
+ * character.
+ */
static void
-insert_thread_id (GHashTable *thread_ids, Xapian::Document doc)
+skip_space_and_comments (const char **str)
{
- string value_string;
- const char *value, *id, *comma;
-
- value_string = doc.get_value (NOTMUCH_VALUE_THREAD);
- value = value_string.c_str();
- if (strlen (value)) {
- id = value;
- while (*id) {
- comma = strchr (id, ',');
- if (comma == NULL)
- comma = id + strlen (id);
- g_hash_table_insert (thread_ids,
- strndup (id, comma - id), NULL);
- id = comma;
- if (*id)
- id++;
+ const char *s;
+
+ s = *str;
+ while (*s && (isspace (*s) || *s == '(')) {
+ while (*s && isspace (*s))
+ s++;
+ if (*s == '(') {
+ int nesting = 1;
+ s++;
+ while (*s && nesting) {
+ if (*s == '(')
+ nesting++;
+ else if (*s == ')')
+ nesting--;
+ else if (*s == '\\')
+ if (*(s+1))
+ s++;
+ s++;
+ }
}
}
+
+ *str = s;
}
-/* Return one or more thread_ids, (as a GPtrArray of strings), for the
- * given message based on looking into the database for any messages
- * referenced in parents, and also for any messages in the database
- * referencing message_id.
- *
- * Caller should free all strings in the array and the array itself,
- * (g_ptr_array_free) when done. */
-static GPtrArray *
-find_thread_ids (Xapian::Database *db,
- GPtrArray *parents,
- const char *message_id)
+/* Parse an RFC 822 message-id, discarding whitespace, any RFC 822
+ * comments, and the '<' and '>' delimeters.
+ *
+ * If not NULL, then *next will be made to point to the first character
+ * not parsed, (possibly pointing to the final '\0' terminator.
+ *
+ * Returns a newly allocated string which the caller should free()
+ * when done with it.
+ *
+ * Returns NULL if there is any error parsing the message-id. */
+static char *
+parse_message_id (const char *message_id, const char **next)
{
- Xapian::PostingIterator child, children_end;
- Xapian::Document doc;
- GHashTable *thread_ids;
- GList *keys, *l;
- unsigned int i;
- const char *parent_message_id;
- GPtrArray *result;
+ const char *s, *end;
+ char *result;
- thread_ids = g_hash_table_new_full (g_str_hash, g_str_equal,
- free, NULL);
+ if (message_id == NULL)
+ return NULL;
- find_messages_by_term (db, "ref", message_id, &child, &children_end);
- for ( ; child != children_end; child++) {
- doc = find_message_by_docid (db, *child);
- insert_thread_id (thread_ids, doc);
- }
+ s = message_id;
- for (i = 0; i < parents->len; i++) {
- parent_message_id = (char *) g_ptr_array_index (parents, i);
- doc = find_message_by_message_id (db, parent_message_id);
- insert_thread_id (thread_ids, doc);
+ skip_space_and_comments (&s);
+
+ /* Skip any unstructured text as well. */
+ while (*s && *s != '<')
+ s++;
+
+ if (*s == '<') {
+ s++;
+ } else {
+ if (next)
+ *next = s;
+ return NULL;
}
- result = g_ptr_array_new ();
+ skip_space_and_comments (&s);
- keys = g_hash_table_get_keys (thread_ids);
- for (l = keys; l; l = l->next) {
- char *id = (char *) l->data;
- g_ptr_array_add (result, id);
+ end = s;
+ while (*end && *end != '>')
+ end++;
+ if (next) {
+ if (*end)
+ *next = end + 1;
+ else
+ *next = end;
}
- g_list_free (keys);
- /* We're done with the hash table, but we've taken the pointers to
- * the allocated strings and put them into our result array, so
- * tell the hash not to free them on its way out. */
- g_hash_table_steal_all (thread_ids);
- g_hash_table_unref (thread_ids);
+ if (end > s && *end == '>')
+ end--;
+ if (end <= s)
+ return NULL;
+
+ result = strndup (s, end - s + 1);
+
+ /* Finally, collapse any whitespace that is within the message-id
+ * itself. */
+ {
+ char *r;
+ int len;
+
+ for (r = result, len = strlen (r); *r; r++, len--)
+ if (*r == ' ' || *r == '\t')
+ memmove (r, r+1, len);
+ }
return result;
}
-/* Add a term for each message-id in the References header of the
- * message. */
+/* Parse a References header value, putting a copy of each referenced
+ * message-id into 'hash'. */
static void
-parse_references (GPtrArray *array,
- const char *refs_str)
+parse_references (GHashTable *hash,
+ const char *refs)
{
- GMimeReferences *refs, *r;
- const char *message_id;
+ char *ref;
- if (refs_str == NULL)
+ if (refs == NULL)
return;
- refs = g_mime_references_decode (refs_str);
+ while (*refs) {
+ ref = parse_message_id (refs, &refs);
- for (r = refs; r; r = r->next) {
- message_id = g_mime_references_get_message_id (r);
- g_ptr_array_add (array, g_strdup (message_id));
+ if (ref)
+ g_hash_table_insert (hash, ref, NULL);
}
+}
- g_mime_references_free (refs);
+char *
+notmuch_database_default_path (void)
+{
+ if (getenv ("NOTMUCH_BASE"))
+ return strdup (getenv ("NOTMUCH_BASE"));
+
+ return g_strdup_printf ("%s/mail", getenv ("HOME"));
}
notmuch_database_t *
notmuch_database_create (const char *path)
{
- char *notmuch_path;
+ notmuch_database_t *notmuch = NULL;
+ char *notmuch_path = NULL;
struct stat st;
int err;
+ char *local_path = NULL;
+
+ if (path == NULL)
+ path = local_path = notmuch_database_default_path ();
err = stat (path, &st);
if (err) {
fprintf (stderr, "Error: Cannot create database at %s: %s.\n",
path, strerror (errno));
- return NULL;
+ goto DONE;
}
if (! S_ISDIR (st.st_mode)) {
fprintf (stderr, "Error: Cannot create database at %s: Not a directory.\n",
path);
- return NULL;
+ goto DONE;
}
notmuch_path = g_strdup_printf ("%s/%s", path, ".notmuch");
if (err) {
fprintf (stderr, "Error: Cannot create directory %s: %s.\n",
notmuch_path, strerror (errno));
- free (notmuch_path);
- return NULL;
+ goto DONE;
}
- free (notmuch_path);
+ notmuch = notmuch_database_open (path);
+
+ DONE:
+ if (notmuch_path)
+ free (notmuch_path);
+ if (local_path)
+ free (local_path);
- return notmuch_database_open (path);
+ return notmuch;
}
notmuch_database_t *
notmuch_database_open (const char *path)
{
- notmuch_database_t *notmuch;
- char *notmuch_path, *xapian_path;
+ notmuch_database_t *notmuch = NULL;
+ char *notmuch_path = NULL, *xapian_path = NULL;
struct stat st;
int err;
+ char *local_path = NULL;
+ unsigned int i;
- g_mime_init (0);
+ if (path == NULL)
+ path = local_path = notmuch_database_default_path ();
notmuch_path = g_strdup_printf ("%s/%s", path, ".notmuch");
err = stat (notmuch_path, &st);
if (err) {
- fprintf (stderr, "Error: Cannot stat %s: %s\n",
- notmuch_path, strerror (err));
- free (notmuch_path);
- return NULL;
+ fprintf (stderr, "Error opening database at %s: %s\n",
+ notmuch_path, strerror (errno));
+ goto DONE;
}
xapian_path = g_strdup_printf ("%s/%s", notmuch_path, "xapian");
- free (notmuch_path);
- /* C++ is so nasty in requiring these casts. I'm almost tempted to
- * write a C wrapper for Xapian... */
- notmuch = (notmuch_database_t *) xmalloc (sizeof (notmuch_database_t));
- notmuch->path = xstrdup (path);
+ notmuch = talloc (NULL, notmuch_database_t);
+ notmuch->path = talloc_strdup (notmuch, path);
try {
notmuch->xapian_db = new Xapian::WritableDatabase (xapian_path,
Xapian::DB_CREATE_OR_OPEN);
+ notmuch->query_parser = new Xapian::QueryParser;
+ notmuch->query_parser->set_default_op (Xapian::Query::OP_AND);
+ notmuch->query_parser->set_database (*notmuch->xapian_db);
+
+ for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++) {
+ prefix_t *prefix = &BOOLEAN_PREFIX_EXTERNAL[i];
+ notmuch->query_parser->add_boolean_prefix (prefix->name,
+ prefix->prefix);
+ }
} catch (const Xapian::Error &error) {
fprintf (stderr, "A Xapian exception occurred: %s\n",
error.get_msg().c_str());
}
- free (xapian_path);
+ DONE:
+ if (local_path)
+ free (local_path);
+ if (notmuch_path)
+ free (notmuch_path);
+ if (xapian_path)
+ free (xapian_path);
return notmuch;
}
void
notmuch_database_close (notmuch_database_t *notmuch)
{
+ delete notmuch->query_parser;
delete notmuch->xapian_db;
- free (notmuch->path);
- free (notmuch);
+ talloc_free (notmuch);
}
const char *
return notmuch->path;
}
+notmuch_private_status_t
+find_timestamp_document (notmuch_database_t *notmuch, const char *db_key,
+ Xapian::Document *doc, unsigned int *doc_id)
+{
+ return find_unique_document (notmuch, "timestamp", db_key, doc, doc_id);
+}
+
+/* We allow the user to use arbitrarily long keys for timestamps,
+ * (they're for filesystem paths after all, which have no limit we
+ * know about). But we have a term-length limit. So if we exceed that,
+ * we'll use the SHA-1 of the user's key as the actual key for
+ * constructing a database term.
+ *
+ * Caution: This function returns a newly allocated string which the
+ * caller should free() when finished.
+ */
+static char *
+timestamp_db_key (const char *key)
+{
+ int term_len = strlen (_find_prefix ("timestamp")) + strlen (key);
+
+ if (term_len > NOTMUCH_TERM_MAX)
+ return notmuch_sha1_of_string (key);
+ else
+ return strdup (key);
+}
+
notmuch_status_t
-notmuch_database_add_message (notmuch_database_t *notmuch,
- const char *filename)
+notmuch_database_set_timestamp (notmuch_database_t *notmuch,
+ const char *key, time_t timestamp)
{
- Xapian::WritableDatabase *db = notmuch->xapian_db;
Xapian::Document doc;
+ unsigned int doc_id;
+ notmuch_private_status_t status;
+ notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
+ char *db_key = NULL;
- GMimeStream *stream;
- GMimeParser *parser;
- GMimeMessage *message;
- GPtrArray *parents, *thread_ids;
+ db_key = timestamp_db_key (key);
- FILE *file;
+ try {
+ status = find_timestamp_document (notmuch, db_key, &doc, &doc_id);
- const char *refs, *in_reply_to;
- const char *message_id;
+ doc.add_value (NOTMUCH_VALUE_TIMESTAMP,
+ Xapian::sortable_serialise (timestamp));
- time_t time;
- unsigned int i;
+ if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND) {
+ char *term = talloc_asprintf (NULL, "%s%s",
+ _find_prefix ("timestamp"), db_key);
+ doc.add_term (term);
+ talloc_free (term);
+
+ notmuch->xapian_db->add_document (doc);
+ } else {
+ notmuch->xapian_db->replace_document (doc_id, doc);
+ }
- file = fopen (filename, "r");
- if (! file) {
- fprintf (stderr, "Error opening %s: %s\n", filename, strerror (errno));
- exit (1);
+ } catch (Xapian::Error &error) {
+ fprintf (stderr, "A Xapian exception occurred: %s.\n",
+ error.get_msg().c_str());
+ ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
}
- stream = g_mime_stream_file_new (file);
+ if (db_key)
+ free (db_key);
+
+ return ret;
+}
- parser = g_mime_parser_new_with_stream (stream);
+time_t
+notmuch_database_get_timestamp (notmuch_database_t *notmuch, const char *key)
+{
+ Xapian::Document doc;
+ unsigned int doc_id;
+ notmuch_private_status_t status;
+ char *db_key = NULL;
+ time_t ret = 0;
- message = g_mime_parser_construct_message (parser);
+ db_key = timestamp_db_key (key);
try {
- doc = Xapian::Document ();
+ status = find_timestamp_document (notmuch, db_key, &doc, &doc_id);
+
+ if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND)
+ goto DONE;
+
+ ret = Xapian::sortable_unserialise (doc.get_value (NOTMUCH_VALUE_TIMESTAMP));
+ } catch (Xapian::Error &error) {
+ goto DONE;
+ }
+
+ DONE:
+ if (db_key)
+ free (db_key);
+
+ return ret;
+}
+
+/* Find the thread ID to which the message with 'message_id' belongs.
+ *
+ * Returns NULL if no message with message ID 'message_id' is in the
+ * database.
+ *
+ * Otherwise, returns a newly talloced string belonging to 'ctx'.
+ */
+const char *
+_resolve_message_id_to_thread_id (notmuch_database_t *notmuch,
+ void *ctx,
+ const char *message_id)
+{
+ notmuch_message_t *message;
+ const char *ret = NULL;
+
+ message = notmuch_database_find_message (notmuch, message_id);
+ if (message == NULL)
+ goto DONE;
+
+ ret = talloc_steal (ctx, notmuch_message_get_thread_id (message));
+
+ DONE:
+ if (message)
+ notmuch_message_destroy (message);
+
+ return ret;
+}
+
+static notmuch_status_t
+_merge_threads (notmuch_database_t *notmuch,
+ const char *winner_thread_id,
+ const char *loser_thread_id)
+{
+ Xapian::PostingIterator loser, loser_end;
+ notmuch_message_t *message = NULL;
+ notmuch_private_status_t private_status;
+ notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
+
+ find_doc_ids (notmuch, "thread", loser_thread_id, &loser, &loser_end);
+
+ for ( ; loser != loser_end; loser++) {
+ message = _notmuch_message_create (notmuch, notmuch,
+ *loser, &private_status);
+ if (message == NULL) {
+ ret = COERCE_STATUS (private_status,
+ "Cannot find document for doc_id from query");
+ goto DONE;
+ }
+
+ _notmuch_message_remove_term (message, "thread", loser_thread_id);
+ _notmuch_message_add_term (message, "thread", winner_thread_id);
+ _notmuch_message_sync (message);
+
+ notmuch_message_destroy (message);
+ message = NULL;
+ }
+
+ DONE:
+ if (message)
+ notmuch_message_destroy (message);
+
+ return ret;
+}
- doc.set_data (filename);
+static notmuch_status_t
+_notmuch_database_link_message_to_parents (notmuch_database_t *notmuch,
+ notmuch_message_t *message,
+ notmuch_message_file_t *message_file,
+ const char **thread_id)
+{
+ GHashTable *parents = NULL;
+ const char *refs, *in_reply_to;
+ GList *l, *keys = NULL;
+ notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
- parents = g_ptr_array_new ();
+ parents = g_hash_table_new_full (g_str_hash, g_str_equal,
+ free, NULL);
- refs = g_mime_object_get_header (GMIME_OBJECT (message), "references");
- parse_references (parents, refs);
+ refs = notmuch_message_file_get_header (message_file, "references");
+ parse_references (parents, refs);
- in_reply_to = g_mime_object_get_header (GMIME_OBJECT (message),
- "in-reply-to");
- parse_references (parents, in_reply_to);
- for (i = 0; i < parents->len; i++)
- add_term (doc, "ref", (char *) g_ptr_array_index (parents, i));
+ in_reply_to = notmuch_message_file_get_header (message_file, "in-reply-to");
+ parse_references (parents, in_reply_to);
- message_id = g_mime_message_get_message_id (message);
+ keys = g_hash_table_get_keys (parents);
+ for (l = keys; l; l = l->next) {
+ char *parent_message_id;
+ const char *parent_thread_id;
- thread_ids = find_thread_ids (db, parents, message_id);
+ parent_message_id = (char *) l->data;
+ parent_thread_id = _resolve_message_id_to_thread_id (notmuch,
+ message,
+ parent_message_id);
- for (i = 0; i < parents->len; i++)
- g_free (g_ptr_array_index (parents, i));
- g_ptr_array_free (parents, TRUE);
- if (message_id) {
- add_term (doc, "msgid", message_id);
- doc.add_value (NOTMUCH_VALUE_MESSAGE_ID, message_id);
+ if (parent_thread_id == NULL) {
+ _notmuch_message_add_term (message, "ref", parent_message_id);
+ } else {
+ if (*thread_id == NULL) {
+ *thread_id = talloc_strdup (message, parent_thread_id);
+ _notmuch_message_add_term (message, "thread", *thread_id);
+ } else if (strcmp (*thread_id, parent_thread_id)) {
+ ret = _merge_threads (notmuch, *thread_id, parent_thread_id);
+ if (ret)
+ goto DONE;
+ }
}
+ }
+
+ DONE:
+ if (keys)
+ g_list_free (keys);
+ if (parents)
+ g_hash_table_unref (parents);
+
+ return ret;
+}
+
+static notmuch_status_t
+_notmuch_database_link_message_to_children (notmuch_database_t *notmuch,
+ notmuch_message_t *message,
+ const char **thread_id)
+{
+ const char *message_id = notmuch_message_get_message_id (message);
+ Xapian::PostingIterator child, children_end;
+ notmuch_message_t *child_message = NULL;
+ const char *child_thread_id;
+ notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
+ notmuch_private_status_t private_status;
+
+ find_doc_ids (notmuch, "ref", message_id, &child, &children_end);
+
+ for ( ; child != children_end; child++) {
+
+ child_message = _notmuch_message_create (message, notmuch,
+ *child, &private_status);
+ if (child_message == NULL) {
+ ret = COERCE_STATUS (private_status,
+ "Cannot find document for doc_id from query");
+ goto DONE;
+ }
+
+ child_thread_id = notmuch_message_get_thread_id (child_message);
+ if (*thread_id == NULL) {
+ *thread_id = talloc_strdup (message, child_thread_id);
+ _notmuch_message_add_term (message, "thread", *thread_id);
+ } else if (strcmp (*thread_id, child_thread_id)) {
+ _notmuch_message_remove_term (child_message, "ref",
+ message_id);
+ _notmuch_message_sync (child_message);
+ ret = _merge_threads (notmuch, *thread_id, child_thread_id);
+ if (ret)
+ goto DONE;
+ }
+
+ notmuch_message_destroy (child_message);
+ child_message = NULL;
+ }
+
+ DONE:
+ if (child_message)
+ notmuch_message_destroy (child_message);
+
+ return ret;
+}
- if (thread_ids->len) {
- unsigned int i;
- GString *thread_id;
- char *id;
+/* Given a (mostly empty) 'message' and its corresponding
+ * 'message_file' link it to existing threads in the database.
+ *
+ * We first looke at 'message_file' and its link-relevant headers
+ * (References and In-Reply-To) for message IDs. We also look in the
+ * database for existing message that reference 'message'.p
+ *
+ * The end result is to call _notmuch_message_add_thread_id with one
+ * or more thread IDs to which this message belongs, (including
+ * generating a new thread ID if necessary if the message doesn't
+ * connect to any existing threads).
+ */
+static notmuch_status_t
+_notmuch_database_link_message (notmuch_database_t *notmuch,
+ notmuch_message_t *message,
+ notmuch_message_file_t *message_file)
+{
+ notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
+ const char *thread_id = NULL;
- for (i = 0; i < thread_ids->len; i++) {
- id = (char *) thread_ids->pdata[i];
- add_term (doc, "thread", id);
- if (i == 0)
- thread_id = g_string_new (id);
- else
- g_string_append_printf (thread_id, ",%s", id);
+ _notmuch_database_link_message_to_parents (notmuch, message,
+ message_file,
+ &thread_id);
- free (id);
+ ret = _notmuch_database_link_message_to_children (notmuch, message,
+ &thread_id);
+
+ if (thread_id == NULL)
+ _notmuch_message_ensure_thread_id (message);
+
+ return ret;
+}
+
+notmuch_status_t
+notmuch_database_add_message (notmuch_database_t *notmuch,
+ const char *filename)
+{
+ notmuch_message_file_t *message_file;
+ notmuch_message_t *message;
+ notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
+
+ const char *date, *header;
+ const char *from, *to, *subject, *old_filename;
+ char *message_id;
+
+ message_file = notmuch_message_file_open (filename);
+ if (message_file == NULL) {
+ ret = NOTMUCH_STATUS_FILE_ERROR;
+ goto DONE;
+ }
+
+ notmuch_message_file_restrict_headers (message_file,
+ "date",
+ "from",
+ "in-reply-to",
+ "message-id",
+ "references",
+ "subject",
+ "to",
+ (char *) NULL);
+
+ try {
+ /* The first order of business is to find/create a message ID. */
+
+ header = notmuch_message_file_get_header (message_file, "message-id");
+ if (header) {
+ message_id = parse_message_id (header, NULL);
+ /* So the header value isn't RFC-compliant, but it's
+ * better than no message-id at all. */
+ if (message_id == NULL)
+ message_id = xstrdup (header);
+ } else {
+ /* No message-id at all, let's generate one by taking a
+ * hash over the file's contents. */
+ char *sha1 = notmuch_sha1_of_file (filename);
+
+ /* If that failed too, something is really wrong. Give up. */
+ if (sha1 == NULL) {
+ ret = NOTMUCH_STATUS_FILE_ERROR;
+ goto DONE;
}
- g_ptr_array_free (thread_ids, TRUE);
- doc.add_value (NOTMUCH_VALUE_THREAD, thread_id->str);
- g_string_free (thread_id, TRUE);
- } else if (message_id) {
- /* If not part of any existing thread, generate a new thread_id. */
- thread_id_t thread_id;
-
- thread_id_generate (&thread_id);
- add_term (doc, "thread", thread_id.str);
- doc.add_value (NOTMUCH_VALUE_THREAD, thread_id.str);
+
+ message_id = g_strdup_printf ("notmuch-sha1-%s", sha1);
+ free (sha1);
+ }
+
+ /* Now that we have a message ID, we get a message object,
+ * (which may or may not reference an existing document in the
+ * database). */
+
+ /* Use NULL for owner since we want to free this locally. */
+ message = _notmuch_message_create_for_message_id (NULL,
+ notmuch,
+ message_id,
+ &ret);
+ free (message_id);
+
+ if (message == NULL)
+ goto DONE;
+
+ /* Has a message previously been added with the same ID? */
+ old_filename = notmuch_message_get_filename (message);
+ if (old_filename && strlen (old_filename)) {
+ ret = NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID;
+ goto DONE;
+ } else {
+ _notmuch_message_set_filename (message, filename);
+ _notmuch_message_add_term (message, "type", "mail");
}
- g_mime_message_get_date (message, &time, NULL);
- doc.add_value (NOTMUCH_VALUE_DATE, Xapian::sortable_serialise (time));
+ ret = _notmuch_database_link_message (notmuch, message, message_file);
+ if (ret)
+ goto DONE;
+
+ date = notmuch_message_file_get_header (message_file, "date");
+ _notmuch_message_set_date (message, date);
- db->add_document (doc);
+ from = notmuch_message_file_get_header (message_file, "from");
+ subject = notmuch_message_file_get_header (message_file, "subject");
+ to = notmuch_message_file_get_header (message_file, "to");
+
+ if (from == NULL &&
+ subject == NULL &&
+ to == NULL)
+ {
+ ret = NOTMUCH_STATUS_FILE_NOT_EMAIL;
+ goto DONE;
+ } else {
+ _notmuch_message_sync (message);
+ }
} catch (const Xapian::Error &error) {
fprintf (stderr, "A Xapian exception occurred: %s.\n",
error.get_msg().c_str());
- return NOTMUCH_STATUS_XAPIAN_EXCEPTION;
+ ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
+ goto DONE;
}
- g_object_unref (message);
- g_object_unref (parser);
- g_object_unref (stream);
+ DONE:
+ if (message)
+ notmuch_message_destroy (message);
+ if (message_file)
+ notmuch_message_file_close (message_file);
- return NOTMUCH_STATUS_SUCCESS;
+ return ret;
}
projects / notmuch / blobdiff